A first study of the inheritance of vision and of the relative influence of heredity and environment on sight / by Amy Barrington and Karl Pearson.

Barrington, Amy.

Date:: 1909

Credit: A first study of the inheritance of vision and of the relative influence of heredity and environment on sight / by Amy Barrington and Karl Pearson. Source: Wellcome Collection.

Provider: This material has been provided by King’s College London. The original may be consulted at King’s College London.

25/74 (page 19)

$Now it must be at once admitted that two factors affect the striking heredity exhibited in these results. The first is the remarkable system of sub-ranges selected by Steiger to tabulate his results. No reasonable explanation is given of this system, and it is utterly unsuited to statistical reduction. The group ranges are occasionally 1*25, occasionally '75, occasionally 1, and again '5. And these ranges vary from table to table, even when one of the pair is the same. It would appear as if some attempt had been made to smooth the frequencies in this manner, possibly in order to get rid of a tendency to read to | instead of ^ diopters, from which Steiger, not- withstanding his criticism of Chauvel (A, p. 18), does not himself seem wholly free. Whatever the source of this choice of sub-ranges *, it enormously increases the labour of the statistician and renders his correction of the moments of doubtful application f. Secondly, there are anomalies of which no explanations are forthcoming. We should expect to find roughly mothers of sons and of daughters to be alike, approximately fathers of sons and of daughters give the same constants; but the mothers of daughters have sensibly higher refraction than mothers of sons. Again, we note that sons of mothers are sensibly more variable than sons of fathers, and daughters of fathers sensibly more variable than daughters of mothers. This increased variability when we deal with the offspring of the opposite sex may be the source of the reduced heredity, in this case between the opposite sexes. It is impossible to say whether this result—important if it were true—is in whole or part produced by the changing of sub-range systems to which we have just referred. We have verified the generally high values of the correlation between parents and offspring by recalculating the coefficients by mean square contingency. Their high values are thus confirmed. The high values of the variabilities for refraction in Table XII compared with those in Table VII, show that we are dealing with highly selected material, but the slight changes of means seem to indicate that the selection is largely one of extremes. The parental variability is raised almost as much as that of the offspring and one hesitates to assert that parents have in the bulk been indirectly selected by selection of their offspring, or that offspring have been indirectly selected by direct selection of their parents. There is nothing in the text to assist us to a conclusion. In the discussion on corneal astigmatism we noticed that at a maximum the correction for selection might reduce the correlation by about 25 °/^. This would reduce our rough value of '61 to 46, a value strikingly close to the average value found for other physical characters in man+. We may safely conclude that corneal refraction is inherited at the same rate as other physical characters in man. Of course the same point impresses itself upon us here as in dealing with astigmatism, the urgent need to measure a large random sample, not an ophthal- mological selection, of the general population of parents and offspring. Such a system of measurements combining astigmatism and refraction would offer splendid material also for testing theories of cross heredity. * We find no justification for it in Steiger's remarks in (B) xvii. p. 445. t Sheppard's correction has been made for an approximate average sub-range. ^ Stature '51, span •48, forearm 42, eye colour -.^0, each based on 4000 to 5000 measurements. 3—2$